Distributed computing, in which many nodes (computers) are caused to perform parallel processing, has been used in recent years. With distributed computing, even if the processing ability of each node is not high, complex processing such as code breaking, medical research, and meteorological analysis can be performed since processing is performed in a distributed manner.
Incidentally, since a network is constructed by many nodes in such distributed computing, there are cases where nodes enter and exit the network. The configuration of the network changes when a node enters or exits, and therefore in this case it is necessary to construct new communication paths, for example, in order to avoid division of the network.
For this reason, techniques have been proposed in which if a node has entered or exited the network, nodes targeted for neither entering nor exiting autonomously construct new communication paths themselves (e.g., see Patent Documents 1 to 3).
Specifically, Patent Documents 1 and 2 disclose systems in which each node is caused to create, in advance, communication paths for the case of an exit, based on the connection information of adjacent nodes. In the systems disclosed in Patent Documents 1 and 2, when a node exits, it notifies the communication paths it created in advance to adjacent nodes, and therefore new communication paths are quickly constructed by the remaining nodes after the exit. For this reason, division of the network is suppressed.
Also, Patent Document 3 discloses a system in which if a fault has occurred in a node in the network, the faulty node is detected by other nodes. In the system disclosed in Patent Document 3, a node connected to the faulty node calculates metric values to adjacent nodes and selects the node with the lowest calculated metric value as the new connection destination. For this reason, even if a fault has occurred, division of the network is suppressed.